Overview

This dataset supports research on masting dynamics in Fagus crenata across northern Japan. It comprises four CSV files covering 102 study sites in the Tohoku region and the Oshima Peninsula of Hokkaido, with qualitative seed production records spanning up to 20 years (1989–2008), along with associated weather cue data derived from meteorological station records. Quantitative seed trap data from the Kanumazawa Riparian Research Forest are also included, together with estimates of latent variables and site-level process model parameters derived from a hierarchical Bayesian population model.

Associated publication:
Masaki T., Hoshizaki K., and Homma C. "Disentangling the effects of geographical location and local climate on masting in Fagus crenata using a comprehensive population model." Basic and Applied Ecology (2025). DOI: 10.1016/j.baae.2025.05.004

Funding:
Japan Society for the Promotion of Science, grants 21H04946 and 24H00783

---

File List

Supplemental climate data: Average climate information for the 102 study sites (WorldClim-derived; CC BY 4.0) is available as a supplemental dataset on Zenodo: DOI: 10.5281/zenodo.20550356

---

Missing Data

Missing values are represented as NA throughout all files.

---

File Descriptions

---

1. List_of_study_sites.csv

Description: Geographic information for 102 Fagus crenata study sites in northern Japan, including 98 sites in the Tohoku region (Honshu) and 4 sites on the Oshima Peninsula (Hokkaido).

Authors of this dataset: Takashi Masaki, Wajiro Suzuki, and Katsuhiro Osumi

Variables:

---

2. Data_used_in_this_study.csv

Description: Qualitative seed production data (seeding ranks) for Fagus crenata at 102 sites over up to 20 years (1989–2008), together with annual weather cue values used in the process model.

Authors of this dataset:

• Tohoku region data (98 sites): Takashi Masaki, Wajiro Suzuki, and Katsuhiro Osumi
• Oshima Peninsula data (4 sites): Kazuhiko Terazawa, Michiyasu Yasaka, Hiromasa Koyama, Hirokazu Kon, Yu Nagasaka, Akiko Nagasaka, Kensuke Onodera, Tomoyuki Abe, and Kazuhiro Minamino

Variables:

Seeding rank key (Obs.seed.rnk):

Weather cue definitions:

• T.HSM (High-temperature summer cue; a single high-temperature pulse in summer): Cumulative positive daily maximum temperature anomaly during July and August, relative to the daily mean for the reference decade 1990–1999. Only continuous runs of positive anomaly days were summed. Original unit: °C·days; standardized to mean = 0 and SD = 1 before use in the model.
• T.LSP (Low-temperature spring cue; a low-temperature pulse in spring): Cumulative negative daily minimum temperature anomaly during April and May, relative to the daily mean for the reference decade 1990–1999. Only continuous runs of negative anomaly days were summed. The resulting values (negative, indicating colder-than-average conditions) were multiplied by −1 and then standardized to mean = 0 and SD = 1 before use in the model as an index of cue intensity. Original unit: °C·days; standardized values are dimensionless.

Weather data source: Meteorological stations located within 20 km of each site, maintained by the Japan Meteorological Agency and accessed via the Ministry of Agriculture, Forestry and Fisheries Research Network, Japan. Site-level temperatures were corrected for elevation difference between the site and the nearest station using a lapse rate of 0.65°C per 100 m.

---

3. Observations_at_Kanumazawa.csv

Description: Annual quantitative measurements of Fagus crenata reproductive output collected from seed traps at the Kanumazawa Riparian Research Forest (39.1°N, 140.9°E), 1990–2008. Data are averages across eight traps located directly beneath mature F. crenata trees within a 1-ha monitoring plot.

Authors of this dataset: Chinatsu Homma, Kazuhiko Hoshizaki, and Takashi Masaki

Trap design: 121 seed traps (0.5 m² each) were installed at 10-m intervals within a 1-ha area. Reproductive organs were collected from May to November each year. Eight traps situated directly beneath mature F. crenata trees were used for the quantitative estimates in this file. Because each female flower contains two ovules, each seed recovered was counted as 0.5 flowers when back-calculating flower density.

Variables:

---

4. Model_estimates.csv

Description: Bayesian posterior estimates of latent variables and site-level process model parameters for each of the 102 study sites. Each row corresponds to one parameter or latent variable for one site. Model structure and estimation procedures are described in detail in the associated publication (DOI: 10.1016/j.baae.2025.05.004).

Authors of this dataset: Takashi Masaki, Kazuhiko Hoshizaki, and Chinatsu Homma

Variables (columns):

Parameter key:

Latent variables (annual, per site):

Process model parameters (per site):

---

Methodological Notes

• Seeding rank assessments for the 98 Tohoku sites were conducted visually every autumn by staff of the Tohoku Regional Forest Office, Forest Agency, Japan, in stands dominated by F. crenata.
• Seeding rank assessments for the 4 Oshima Peninsula sites were derived from seed trap seedfall density data using the same four-level classification criteria.
• Original Tohoku seeding rank data are publicly available (in Japanese) at: https://www.ffpri.affrc.go.jp/labs/tanedas/

---

Sharing and Access Information

This dataset is published under the CC0 1.0 Universal Public Domain Dedication. Supplemental climate data for the study sites (WorldClim-derived; CC BY 4.0) are available on Zenodo (DOI: 10.5281/zenodo.20550356). Users of the climate data should cite the WorldClim database accordingly:

Hijmans, R.J., Cameron, S.E., Parra, J.L., Jones, P.G., and Jarvis, A. (2005). Very high resolution interpolated climate surfaces for global land areas. International Journal of Climatology, 25, 1965–1978. DOI: 10.1002/joc.1276